Do People Use ChatGPT to Find Local Businesses? What the Data Shows

Yes — people use ChatGPT and other AI tools to find local businesses, and the behavior is growing fast from a base that's already commercially meaningful. ChatGPT serves roughly 900 million weekly users, Google's AI Overviews now sit on about 30% of searches (74% of problem-solving queries), and Gartner projects 25% of organic search traffic shifting to AI chatbots and virtual agents. The honest nuance: AI hasn't replaced Google Maps for "plumber near me" — it's absorbing the deciding part of local discovery, where shortlists form.

We watch this from an unusual seat: our software, Radar, observes real AI answers about local businesses daily. So rather than argue from vibes in either direction, here's what the data supports, which local queries moved to AI first, and what a local business should actually do about it.

What the numbers say

The headline stats stack into one picture of changed behavior:

• 68% of US Google searches end without a click (SparkToro, June 2026) — the majority of search behavior now resolves in the answer layer, before any website visit.
• AI Overviews appear on ~30% of Google queries, concentrated on question-form searches (Authoritas, 2025). Google's ~90% search share means this alone rewired a huge slice of local discovery.
• ChatGPT's ~900M weekly users ask it conversational questions traditional search never saw — and when the question is "who should I hire," the answer names specific businesses.
• In our own scan data across local markets, the same buyer questions asked of ChatGPT, Gemini, and Perplexity return concentrated recommendations: typically two to four business names per answer, with one or two businesses dominating repeated runs of a market's questions.

The direction is unambiguous even where precise adoption numbers are noisy: local business discovery is migrating from "scan a list of links" to "read a synthesized recommendation."

Which local searches moved to AI first

Not all local queries migrated equally. From what we observe:

1. Research and comparison questions moved first. "How much does a kitchen remodel cost?", "what's the difference between a crown and a veneer?", "best family lawyer for a custody case" — the anxious homework people do before contacting anyone. These are exactly the queries AI Overviews trigger on and users bring to ChatGPT.
2. High-consideration shortlisting is moving now. "Who are the most reliable roofers in Sarasota — I have a tile roof and two estimates already" is a real pattern: conversational, contextual, unanswerable by a map pin. AI tools like ChatGPT excel here, and this is where recommendations directly reassign revenue.
3. Urgent transactional lookups moved least. "Emergency plumber open now" still runs through Google's local pack and Maps — proximity and hours win, and Google's AI layer is only beginning to reshape those results.

The commercial punchline: the queries that moved first are the ones where decisions form. A homeowner who asked ChatGPT three research questions and got the same company named twice arrives at the "near me" search already leaning.

Demographics and the trust shift

Younger users increasingly treat AI chat as the default starting point — the same cohort shift that once moved local discovery from phone books to search engines, then from desktop to mobile. Two behaviors compound it: voice assistants piping through the same AI models, and AI answers being trusted as recommendations rather than evaluated as ads. When an assistant says "customers praise their communication and cleanup," it reads like advice from a knowledgeable friend — sourced, invisibly, from reviews and web data the business may never have checked.

What this means for a local business (without the panic)

The right response isn't abandoning local SEO — it's recognizing that a second discovery surface now runs on mostly the same inputs:

• Your Google Business Profile and reviews feed both. The local pack reads them; so do the AI systems synthesizing recommendations.
• Your website's clarity feeds both. Pages that answer real questions in extractable language rank in traditional search and supply AI answers. Structured data (schema markup) makes the facts unambiguous for each.
• The new work is measurement. You can see your Google rankings; you cannot see the thousands of AI conversations recommending someone in your market — unless you test. Start with what AI currently says about your business, then the playbook for getting recommended.

Early-mover math still applies: most local competitors haven't checked once. Being the business that's accurate, extractable, and corroborated while rivals ignore the surface is a cheap advantage that won't stay cheap.

Frequently asked questions

How many people use ChatGPT to find local businesses?

No public dataset isolates local-intent ChatGPT queries precisely — what's measurable is scale and direction: ~900M weekly ChatGPT users, AI Overviews on ~30% of Google searches, 68% of searches ending clickless, and Gartner's projected 25% organic-traffic shift to AI agents. Directionally: a minority of local discovery today, growing quarter over quarter, concentrated in research and shortlisting queries.

Do people trust AI recommendations for local services?

Increasingly, yes — because AI answers read as synthesized advice rather than advertising, and they justify picks with review themes ("praised for fast communication"). That trust transfers to whoever gets named, which is exactly why being absent from — or misdescribed in — those answers costs real revenue.

Will AI replace Google for local search?

Replace, no; restructure, yes. Google Maps and the local pack still dominate urgent transactional lookups, and Google's own AI Overviews keep discovery inside Google. What changes is where shortlists form: research and comparison questions increasingly resolve in AI answers before anyone types "near me."

How do I know if AI sends me customers?

Two layers: ask new customers how they found you (AI referrals routinely get self-reported as "I searched online"), and measure the answer layer directly — test your market's buyer questions across ChatGPT, Gemini, and Perplexity and record who gets named. Our free scan automates that with receipts per answer.